R Notebook

Tidy Tuesday Week 20

Internet Access, US Broadband, data from Microsoft by way of The Verge.

TidyTuesday week 20 | Data from Microsoft and The Verge

Data definition

variable	class	description
st	character	2 letter abbreviation of states in the United States https://www.iso.org/obp/ui/#iso:code:3166:US
county_id	double	4 to 5 digit code used to represent the county (last 3 digits) and the state (first digit or first 2 digits) https://www.census.gov/geographies/reference-files.html
county_name	character	County Name
broadband_availability_per_fcc	character	percent of people per county with access to fixed terrestrial broadband at speeds of 25 Mbps/3 Mbps as of the end of 2017 https://www.fcc.gov/document/broadband-deployment-report-digital-divide-narrowing-substantially-0
broadband_usage	character	percent of people per county that use the internet at broadband speeds based on the methodology explained above. Data is from November 2019.

Visualization

# load libaries 
library(tidytuesdayR)
library(tidyverse)
library(urbnmapr)
library(urbnthemes)
library(janitor)
library(colorspace)
library(wesanderson)
library(patchwork)
library(ggstatsplot)

# import data
broadband <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2021/2021-05-11/broadband.csv')

── Column specification ──────────────────────────────────────────────────────────────────────────────
cols(
  ST = col_character(),
  `COUNTY ID` = col_double(),
  `COUNTY NAME` = col_character(),
  `BROADBAND AVAILABILITY PER FCC` = col_character(),
  `BROADBAND USAGE` = col_character()
)

# clean data 
# reference: Jesse Mostipak  
broadband = broadband %>% janitor::clean_names()
glimpse(broadband)

Rows: 3,143
Columns: 5
$ st                             <chr> "AL", "AL", "AL", "AL", "AL", "AL", "AL", "AL", "AL", "AL", "…
$ county_id                      <dbl> 1001, 1003, 1005, 1007, 1009, 1011, 1013, 1015, 1017, 1019, 1…
$ county_name                    <chr> "Autauga County", "Baldwin County", "Barbour County", "Bibb C…
$ broadband_availability_per_fcc <chr> "0.81", "0.88", "0.59", "0.29", "0.69", "0.06", "0.78", "0.93…
$ broadband_usage                <chr> "0.28", "0.30", "0.18", "0.07", "0.09", "0.05", "0.11", "0.32…

broadband = broadband %>% mutate(broadband_usage= as.numeric(broadband_usage),
                                 broadband_availability_per_fcc = as.numeric(broadband_availability_per_fcc))

NAs introduced by coercionNAs introduced by coercion

# us counties shp 
counties_sf <- get_urbn_map("counties", sf = TRUE)

U.S. Availability and Usage by County

# availability by county
p1 = counties_sf %>% 
  left_join(broadband, by="county_name") %>%
  #filter(st=="CA") %>%
  ggplot() + 
  geom_sf(mapping = aes(fill = broadband_availability_per_fcc),
          color = "#ffffff", size = 0.25) +
  coord_sf(datum = NA) + 
  scale_fill_gradientn(labels = scales::percent,
                       colours = wes_palette("Zissou1", 100, type = "continuous"), 
                       na.value="grey89",
                       trans="reverse") +
  theme_minimal(base_size = 10) + 
  theme(plot.subtitle=element_text(size=9),
        plot.title=element_text(size=14)) + 
  guides(fill = guide_colorbar(title.position = "top", 
                                title.hjust = .5, 
                                reverse=TRUE,
                                barheight = unit(10, "lines"), 
                                barwidth = unit(.5, "lines"))) +
  labs(fill="",
       title="U.S. Broadband Availability by County",
       subtitle="Percent of people per county with access to fixed terrestrial broadband at speeds of 25 Mbps/3 Mbps, as of the end of 2017")

# usage by county
p2 = counties_sf %>% 
  left_join(broadband, by="county_name") %>%
  #filter(st=="CA") %>%
  ggplot() + 
  geom_sf(mapping = aes(fill = broadband_usage),
          color = "#ffffff", size = 0.25) +
  coord_sf(datum = NA) + 
  scale_fill_gradientn(labels = scales::percent,
                       colours = wes_palette("Zissou1", 100, type = "continuous"), 
                       na.value="grey89",
                       trans="reverse") +
  theme_minimal(base_size = 10) + 
  theme(plot.subtitle=element_text(size=9),
        plot.title=element_text(size=14)) + 
  guides(fill = guide_colorbar(title.position = "top", 
                                title.hjust = .5, 
                                reverse=TRUE,
                                barheight = unit(10, "lines"), 
                                barwidth = unit(.5, "lines"))) +
  labs(fill="",
       title="U.S. Broadband Usage by County",
       subtitle="Percent of people per county that use the internet at broadband speeds of 25 Mbps/3 Mbps, as of November 2019")

p1/ p2 + plot_annotation(caption='TidyTuesday week 20 | Data from The Microsoft') & theme(plot.caption = element_text(hjust=0.5))

Texas: Broadband Availability and Usage

• shared on Twitter

# availability Texas
plot1 = counties_sf %>% filter(state_abbv=="TX") %>%
  left_join(broadband, by="county_name") %>%
  ggplot() + 
  geom_sf(mapping = aes(fill = broadband_availability_per_fcc),
          color = "#ffffff", size = 0.25) +
  coord_sf(datum = NA) + 
  scale_fill_gradientn(labels = scales::percent,
                       colours = wes_palette("Zissou1", 100, type = "continuous"), 
                       na.value="grey89",
                       trans="reverse") +
  theme_minimal(base_size = 10) + 
  theme(plot.subtitle=element_text(size=9),
        plot.title=element_text(size=14)) + 
  guides(fill = guide_colorbar(title.position = "top", 
                                title.hjust = .5, 
                                reverse=TRUE,
                                barheight = unit(10, "lines"), 
                                barwidth = unit(.5, "lines"))) +
  labs(fill="",
       title="Broadband Availability In Texas, U.S.",
       subtitle="Percent of people per county with access to fixed terrestrial broadband at speeds\nof 25 Mbps/3 Mbps, as of the end of 2017")

# Usage Texas
plot2 = counties_sf %>% filter(state_abbv=="TX") %>%
  left_join(broadband, by="county_name") %>%
  ggplot() + 
  geom_sf(mapping = aes(fill = broadband_usage),
          color = "#ffffff", size = 0.25) +
  coord_sf(datum = NA) + 
  scale_fill_gradientn(labels = scales::percent,
                       colours = wes_palette("Zissou1", 100, type = "continuous"), 
                       na.value="grey89",
                       trans="reverse") +
  theme_minimal(base_size = 10) + 
  theme(plot.subtitle=element_text(size=9),
        plot.title=element_text(size=14)) + 
  guides(fill = guide_colorbar(title.position = "top", 
                                title.hjust = .5, 
                                reverse=TRUE,
                                barheight = unit(10, "lines"), 
                                barwidth = unit(.5, "lines"))) +
  labs(fill="",
       title="Broadband Usage In Texas, U.S.",
       subtitle="Percent of people per county that use the internet at broadband speeds of\n25 Mbps/3 Mbps, as of November 2019")

plot1/ plot2 + plot_annotation(caption='TidyTuesday week 20 | Data from Microsoft') & theme(plot.caption = element_text(hjust=0.5))

Broadband Availability by State

broadband %>% drop_na() %>% group_by(st) %>%
  summarise(med = median(broadband_usage)) -> table_median

# boxplot of broadband usage by state
broadband %>% drop_na() %>%
  ggplot(aes(x=reorder(st,broadband_usage), y=broadband_usage)) + 
  geom_boxplot(color="grey60") + 
  geom_point(data=table_median, aes(x= st, y=med, fill=med), shape=23, color="slategrey",show.legend=F) +
  scale_fill_continuous_sequential(palette="Viridis") +
  scale_y_continuous(label=scales::percent) +
  theme_minimal(base_size=10) + 
  theme(panel.grid.minor=element_blank()) +
  coord_flip()

# usage 5 states

# geom_flat_violin function from: https://gist.github.com/dgrtwo/eb7750e74997891d7c20
# reference: https://neuroconscience.wordpress.com/2018/03/15/introducing-raincloud-plots/
# reference: https://bitsandgen.es/my-first-shiny-app-raincloudplots/

pop = c("CA","TX","FL","NY", "PA") # states for plot
broadband_pop = broadband %>% filter(st %in% pop)

broadband_pop %>%
  ggplot(aes(x=fct_rev(st), 
             y=broadband_usage, fill=st, color=st)) + 
  geom_point(position=position_jitter(0.15),size=1.5, alpha=0.6,shape=as.numeric(16)) + 
  geom_flat_violin(position= position_nudge(x=0.25, y=0),adjust=2,alpha=0.9, trim='TRUE',scale='width') +
  geom_boxplot(aes(x=st,y=broadband_usage,fill=st), position=position_nudge(x=0.25, y=0),
               width=0.1, outlier.shape = NA, varwidth=FALSE, color="black", alpha=0.3) + 
  scale_y_continuous(label=scales::percent) +
  scale_fill_manual(values=c("#588b8b","#f28f3b","#457b9d","#c8553d","#5b5f97"))+
  scale_color_manual(values=c("#588b8b","#f28f3b","#457b9d","#c8553d","#5b5f97")) +
  coord_flip() + 
  theme_minimal(base_size = 10) +
  theme(legend.position="none",
        plot.title.position = "plot",
        plot.subtitle=element_text(face="bold", size=10),
        axis.title=element_text(size=9, face="bold"),
        plot.margin=ggplot2::margin(1,1,1,1,"cm")) +
  expand_limits(x = 5.25) +
  labs(x="State",y="Broadband Usage",
       subtitle="Comparision of broadband usage across five U.S. States\n")

# comparing five states

# availability 
s1 = ggstatsplot::ggbetweenstats(
  data=broadband_pop, 
               x= st, y=broadband_availability_per_fcc,
               title="Distribution of broadband availability",
               package="wesanderson",
               palette="Darjeeling1") +
  ggplot2::theme(plot.margin=ggplot2::margin(1,1,1,1,"cm"))

# usage 
s2 = ggstatsplot::ggbetweenstats(
  data=broadband_pop, 
               x= st, y=broadband_usage,
               title="Distribution of broadband usage",
               package="wesanderson",
               palette="Darjeeling1") +
  ggplot2::theme(plot.margin=ggplot2::margin(1,1,1,1,"cm"))

s1 | s2